![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
Members of the same household share similar social attitudes, but the source of the similarity in attitudes may be attributed to many processes. This study uses data from a randomized field experiment to identify contagion in attitude change about anti-transgender prejudice. During a face-to-face canvassing experiment, registered voters who answered the door were exposed to either a message encouraging active perspective taking intended to reduce transphobia or a recycling message. Here, I show that the messages delivered to one household member are likely to reduce anti-transgender prejudice in the cohabitants as well. This finding suggests that door-to-door canvassing messages intended to elicit attitude change are likely to be socially transmitted.
Notes
1. Support transgender teacher, transgender bathroom use morally wrong, and gender norms about dress were asked on the 3-week and 6-week surveys only.
2. The use of factor analysis should decrease measurement error from these measures, which is particularly important given that they have not been validated. For the purposes of this study, the factor analysis creates index measures related to attitudes toward transgender people while reducing measurement error, but the factor variable may or may not represent a latent attitude toward transgender people.
3. The covariates from the voter file are age, an indicator for whether the participant is black, an indicator for whether the participant is Hispanic, an indicator for whether the participant is a registered Democrat, and gender. The covariates from the baseline survey are Support for transgender protection law one, Support for transgender protection law two, Feeling thermometer toward transgender people, Transgender morally wrong one, Transgender morally wrong two, Support friend’s sex change, a feeling thermometer score for Barack Obama, a feeling thermometer score for gay people, the participant’s response to a 7-point political ideology question, the participant’s response to a 7-point political party identification question, whether the participant had met a gay or lesbian person, whether the participant had med a transgender person, frequency of attending religious services, support for same sex marriage, social dominance orientation scale score, and the language in which the participant chose to take the survey (English or Spanish).
4. By splitting the sample, I am investigating whether or not a treatment effect exists within a subsample rather than whether the treatment effect varies across groups. Previous work has similarly split samples based on dyad characteristics to investigate peer influence (Bond et al., Citation2012; Jones et al., Citation2017; Fowler & Christakis, Citation2009; Shakya et al., Citation2016). I provide a direct test of the difference in treatment effect in the supplementary materials in which I treat age difference as a moderator.